Today’s technologies, especially AI, offer great opportunities for optimisation and improvement of many court processes, where many of the tasks are repetitive and human intervention does not add significant value. Enforcement trials, due to their standardised nature, are ideal for the implementation of AI solutions. Moreover, there are already experiences in Argentina of automating enforcement trials that take advantage of the benefits that technology has to offer.
Illustration: Papanika.
This work is the result of a collaboration between Fundar and IALAB, the Innovation and Artificial Intelligence Laboratory of the Law School of the University of Buenos Aires.
It is aimed at court personnel from all over the country and at the technology areas of judiciaries and magistrates’ councils that wish to improve efficiency in judicial processes with the support of technological tools (and, in particular, those based on AI) as part of their strategic plans.
Use of AI in judicial proceedings
What are we talking about when we talk about using AI to improve justice?
The application of various AI tools to improve judicial processes can result in better use of available resources, valuing human work and enhancing human intervention in complex cases.
For example, the implementation of automated case file management processes can significantly reduce processing times and minimise human error, while the use of algorithms can be used for the detection and evaluation of enforceable titles by analysing documents using computer vision and natural language processing.
Why enforcement trials?
Enforcement trials constitute a significant part of the total caseload in civil and commercial courts of first instance across the country and generally involve a considerable burden of administrative tasks. They usually involve a considerable burden of administrative tasks. The automation of these tasks would free up valuable resources for more complex cases and speed up the resolution of trials, directly benefiting citizens. In other words, less processing time and optimisation of human resources.
AI systems can perform different tasks depending on the needs and requirements of each judicial process. Enforcement proceedings are great candidates for automation through AI because of their structured, repetitive, and non-contentious nature. Most proceedings are conducted without opposition or complex legal discussions that require in-depth analysis of the facts or applicable law.
Enforcement proceedings are initiated on the basis of a title – debt note, cheque, promissory note or other – so a first dimension for the automation of enforcement proceedings would be the automated generation of the title, so that AI solutions are not a stopgap.
AI capabilities for Enforcement trials
| Task | Description |
| Data extraction from free-text documents | The AI-based system can extract specific data requested from unstructured documents, such as lawsuits or other court documents. |
| Data verification | The AI system can estimate with some degree of accuracy the existence of specific data or the occurrence of specific assumptions. It can also give an estimate of the consistency of data, or measure the difference from different sources. |
| Integration of generative AI tools with management systems | AI tools can be integrated into court records management systems, allowing flexible adaptation to existing software. |
| Document writing assistance | LLM-based systems can be used to generate base templates or draft texts for later use, while a human verifies and customises them for their particular purposes. |
| Document analysis | Generative AI could be asked to analyse and extract relevant information from normative texts, doctrine or jurisprudence for use in drafting documents. |
| Database development | Generative AI can extract and process data to form databases. This is especially useful in the initial phase of enforcement processes, where variable data can be extracted from demands for later use. |
Three experiences of automating enforcement trials
Experience 1
Prometea in Chaco Province
The implementation of Prometea in the Court of Enforcement Proceedings, Bankruptcy and Insolvency of the 2nd District of the Province of Chaco allowed the automation of the generation of payment orders, fee and interest calculations, and the transcription of amounts to text, tasks that were previously performed manually.
The adaptation of Prometea to this process also simplified document review through a system based on decision trees, reducing the time needed to project judgements from hours to minutes and allowing court staff to concentrate on more complex tasks.
With Prometea, the average time between the filing of the lawsuit and the issuance of the monitoring judgment has been shortened considerably: cases are resolved in just three days, in contrast to the month-long time taken by other similar courts.
Experience 2
Generative AI in Rio Negro Province
In the Province of Río Negro, the automation of enformcement trials through generative AI has been implemented for tax enforcement processes. The system automates everything from the formal control of the claim and the debt bill, to text conversion, data validation and the generation of judgments.
It uses an OpenAI API to validate data and generates responses in JSON format, which are automatically integrated into the electronic case file management system.
The system implemented in the Province of Río Negro has shown remarkable results, allowing judgments to be issued in record time, as little as one hour from the filing of the lawsuit.
Experience 3
Generative AI in Chac Province
The prototype developed at the Judicial Management Office of the Chaco Peace Courts uses generative AI to automate the management of fine and patent enforcement. This system classifies types of executions, extracts and verifies entities in court documents, and generates draft resolutions.
